Abstract
Radiographic signs in 64 cats that had radiography as part of the diagnostic work-up for suspected nasal disease were reviewed in a blinded fashion. Final diagnoses in these cats were rhinitis in 27, primary nasal neoplasia in 21 and non-nasal disease in 16. The signs with highest predictive value for nasal neoplasia were displacement of midline structures (73%), unilateral generalised soft tissue opacity (70%), unilateral generalised loss of turbinate detail (69%) and evidence of bone invasion (64%). The only radiographic finding that occurred more frequently in cats with rhinitis was a nasal cavity within normal limits, and the predictive value of this sign was only 38%. Radiographic signs in cats with nasal neoplasia are similar to those reported in dogs, whereas the radiographic signs in cats with rhinitis are variable and non-specific, and may be absent.
Introduction
Nasal diseases of cats may be divided predominantly into inflammatory conditions and neoplasms (Bright 1981). Feline rhinitis may be caused by various respiratory infections including calicivirus, feline herpesvirus-1, Bordetella bronchiseptica, Chamydilophila felis and Cryptococcus neoformans (Bradley 1984, Cape 1992, Malik et al 1992, Ford and Levy 1994, Van Pelt and Lappin 1994, Binns et al 1999, Mochizuki et al 2000). Rhinitis has also beenassociated with systemic conditions, such as feline immunodeficiency virus infection (Hopper et al 1989). Non-infectious feline rhinitides may occur as a result of eosinophilic or lymphoplasmacytic infiltration or foreign body (Bright 1981). Rhinitis (and any secondary sinusitis) may become chronic when inflammation is severe enough to cause structural changes in the nasal conchae that compromise their immunological function (Bright 1981, Bradley 1984, Levy and Ford 1994). Feline nasal neoplasms include a wide variety of cell types with malignant epithelial neoplasms and lymphoma predominating (Levy and Ford 1994). In a recent review of 123 feline nasal and sinus neoplasms, the most frequent diagnoses were lymphoma (28%), adenocarcinoma (15%) and squamous cell carcinoma (14%)(Mukaratirwa et al 2001). Inflammatory polyps, which often affect the nasopharynx and may obstruct the choanae, must be distinguished from neoplasms (Bradley 1984, Kapatkin et al 1990, Allen et al 1999).
Clinical signs in cats with nasal and nasopharyngeal diseases often include inappetance, nasal discharge, sneezing, stertor and mouth breathing. Epistaxis, facial deformity and exophthalmos are additional signs that are usually associated with relatively aggressive conditions (Levy and Ford 1994, Van Pelt and Lappin 1994). For cats with short-lived, serous nasal discharge suggestive of viral rhinitis, diagnosis may be based on clinical signs (Ford and Levy 1994). Viral isolation to identify carriers is usually reserved for management of recurrent outbreaks of rhinitis in multi-cat households or for epidemiological studies (Ford and Levy 1994, Mochizuki et al 2000). When clinical signs become chronic or more severe or nasal discharge becomes purulent, further diagnostic work-up is indicated (Cape 1992, Levy and Ford 1994).
Radiography is one of the principal diagnostic methods used in the investigation of cats (and dogs) with chronic nasal signs; however, there is relatively little published information about radiographic signs in cats with nasal disease compared to dogs. Coulson (1988) and Farrow (1994) have reviewed this subject. In 10 cats with chronic rhinosinusitis radiographed by Cape (1992), there was fluid/soft tissue opacity in the nasal cavity in seven, loss of turbinate detail in four, fluid/soft tissue was present in the ipsilateral frontal sinus in four, signs of destruction of the nasal or vomer bones were identified in two, and two cats had no radiographic signs. Cox et al (1991) briefly described the radiographic signs in 13 cats with nasal or sinus neoplasms. In 12/13 cats there was increased opacity affecting intranasal structures and in five there was evidence of bone destruction. Radiographic signs in a series of 29 cats, including 18 with nasal neoplasia and 11 with rhinitis were reported by O'Brien et al (1996). Unilateral loss of turbinates, bone erosion and tooth loss were associated with nasal neoplasia (O'Brien et al 1996). The prevalence of various other signs, including facial swelling and deviation or lysis of midline nasal structures, was similar in cats with neoplasia and chronic rhinitis (O'Brien et al 1996).
On the basis of these reports, there appears to be considerable overlap in the radiographic signs that may be observed in these conditions, hence radiography may be relatively less useful as a means of distinguishing rhinitis from nasal neoplasia in cats than it is in dogs. The aim of the present study was to describe the radiographic signs in a series of cats with nasal disease and to identify any signs that might be used to support a specific diagnosis.
Materials and methods
Medical records at the Queen Mother Hospital for Animals were searched for cats that had radiography of the skull and a nasal examination that included endoscopy. In each instance skull radiographs included a dorsoventral (intraoral) projection. Cases were collected by one of the investigators (SR) who did not participate in the radiographic evaluation. Medical records were reviewed with respect to age, sex, breed, and presence of clinical signs referable to conditions affecting the nasal cavity (nasal discharge, epistaxis, sneezing, facial deformity, exophthalmos, ocular discharge, stertor and dyspnoea). Cats were categorised into those with rhinitis, those with nasal neoplasia and those with a non-nasal condition. Diagnosis of rhinitis was based on presence of inflammatory cells (and absence of neoplastic cells) in nasal flushings or endoscopic biopsies, positive aerobic bacterial culture and/or retrieval of a nasal foreign body. In recognition of the fact that neoplasia may be missed by cytological or histological examination, some cats categorised as rhinitis in this study had multiple examinations involving endoscopy, all of which were negative for neoplasia. Also follow-up by telephone contact with the referring veterinary surgeons was used selectively to check that no cats categorised as rhinitis had a progressive course or additional test results that suggested underlying neoplasia. Diagnosis of nasal neoplasia was based on histological examination of nasal biopsy. All cats classified as having non-nasal conditions had no abnormalities identified by nasal endoscopy.
Radiographs were examined without knowledge of any clinical information by two radiologists (CRL, PM), who reached a consensus about the radiographic signs in each case and recorded their observations using a custom-designed form. The following signs were assessed: lesion location (unilateral, bilateral, rostral, middle or caudal part of nasal cavity or generalised), loss of turbinate detail (generalised, focal, multifocal), soft tissue/fluid opacities (generalised, focal, multifocal), lucent foci (generalised, focal, multifocal), presence of abnormal intranasal calcification, displacement of midline structures, evidence of invasion of bones surrounding the nasal cavity, evidence of facial deformity, lesions affecting the frontal sinuses (fluid/soft tissue content, aggressive or non-aggressive bone lesions) and dental lesions (absent teeth, loss of lamina dura, erosion of teeth).
On the basis of the radiographic signs, cats were also assigned a score based on the scheme of nasal radiographic patterns described by Myer (1998): “1, normal radiographic appearance of both nasal passages; 2, areas of increased soft tissue opacity superimposed over normal conchal pattern; 3, areas of increased soft tissue opacity superimposed over areas of conchal destruction; 4, areas of decreased opacity owing to conchal destruction without accompanying soft tissue opacity; 5, a mixed pattern with areas of conchal destruction and superimposed soft tissue opacity interspersed with areas of conchal destruction alone”.
The median ages of cats with rhinitis and nasal neoplasia were compared using the Mann–Whitney U-test. Prevalence of clinical and radiographic signs in cats with nasal neoplasia and rhinitis were compared using Fisher's exact test.
Results
Sixty-four cats satisfied the criteria for inclusion, including 27 with rhinitis, 21 with primary nasal neoplasia and 16 cats without nasal disease(Table 1). Rhinitis was associated with bacterial infection in 14 cats. Of these, there was evidence of a single bacterial infection in nine cats: Pasteurella multocida in six, Staphylococcus intermedius in one, Pseudomonas aeruginosa in one and B bronchiseptica in one. Five cats had mixed bacterial infection. The majority of nasal neoplasms were classified as carcinomata.
Summary of diagnoses in 64 cats having nasal radiography
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table1.png)
The age of cats ranged from 5 months to 17 years. A variety of breeds were represented, including 36 domestic shorthairs, five domestic longhairs, seven Siamese, five Burmese, four Persian, three Russian Blue, two Somali, one British shorthair and an Egyptian. Cats with nasal neoplasia were significantly older than cats with rhinitis (Table 2). The proportion of the cats with rhinitis that were pure-bred (44%) was significantly greater than the proportion of cats with nasal neoplasia that were pure-bred (19%). Similar clinical signs were recorded in cats with either rhinitis or nasal neoplasia (Table 3), although significantly more cats with rhinitis sneezed and significantly more cats with nasal neoplasia had epistaxis and/or facial deformity. The median duration of clinical signs prior to referral of cats was 3 months both for cats with rhinitis and those with nasal neoplasia.
Signalment of 64 cats having nasal radiography
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table2.png)
Key: M, male; F, female.
Cats with rhinitis were significantly younger than cats with nasal neoplasia (p=0.001).
The proportion of the cats with rhinitis that were pure-bred was significantly greater than the proportion of cats with nasal neoplasia that were pure-bred (p=0.05).
Prevalence of selected clinical signs in 48 cats with rhinitis or nasal neoplasia
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table3.png)
NS, difference not statistically significant.
Nasal structures were interpreted as abnormal radiographically in 19/27 (70%) cats with rhinitis, 21/21 (100%) cats with nasal neoplasia, and 3/16 (19%) cats with non-nasal disease. The radiographic signs recorded in the cats with rhinitis or nasal neoplasia are summarised in Table 4 and examples illustrated in Figs 1 and 2. The only radiographic finding that occurred more frequently in cats with rhinitis was a nasal cavity within normal limits (p=0.006). The predictive value of this sign was only 38%. Signs that occurred more frequently in cats with nasal neoplasia were displacement of midline structures, unilateral generalised soft tissue opacity, unilateral generalised loss of turbinate detail, displacement of midline structures and evidence of bone invasion (p<0.05). The signs with highest predictive value for nasal neoplasia were displacement of midline structures (73%), unilateral generalised soft tissue opacity (70%), unilateral generalised loss of turbinate detail (69%) and evidence of bone invasion (64%) (Table 5). Dental abnormalities were observed frequently in all groups of cats.
Summary of radiographic signs identified in 48 cats with rhinitis or nasal neoplasia
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table4.png)
NS, difference not statistically significant.
Predictive values of radiographic signs in 64 cats having nasal radiography
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table5.png)
Predictive value of sign for disease 00023-8/unzip/10.1016_S1098-612X(03)00023-8-img1.png)
00023-8/unzip/10.1016_S1098-612X(03)00023-8-fig1.png)
Examples of radiographic signs in cats with chronic rhinitis. (A) Dorsoventral (intraoral) radiograph of a cat with rhinitis associated with a mixed bacterial infection, in which the intranasal structures were interpreted as within normal limits; (B) dorsoventral (intraoral) radiograph of a cat with chronic necrotising rhinitis, in which there is a generalised increase in opacity with loss of turbinate detail affecting the right nasal cavity; (C) dorsoventral (intraoral) radiograph of a cat with an airgun pellet lodged in the rostral part of the left nasal cavity. There are lucent foci (arrow) compatible with localised turbinate destruction just caudal to the foreign body.
00023-8/unzip/10.1016_S1098-612X(03)00023-8-fig2.png)
Examples of radiographic signs in cats with nasal neoplasia. (A) Dorsoventral (intraoral) radiograph of a cat with a granular cell tumour originating in the right nasal cavity. There is a relatively uniform increased opacity with loss of turbinate detail affecting the right nasal cavity and focal lucent area (∗) as a result of bone destruction; (B) dorsoventral (intraoral) radiograph of a cat with a poorly differentiated sarcoma of the left nasal cavity. There is a partially calcified mass (∗) that extends across the midline (arrowheads) and laterally (arrow). Multiple teeth are absent as a result of pre-existing dental disease; (C) dorsoventral (intraoral) radiograph of a cat with a carcinoma, in which there is a bilateral, relatively uniform increase in opacity and loss of turbinate detail; (D) lateral radiograph of a cat with nasal carcinoma, in which there is evidence of destruction of the cribriform plate (arrow) as a result of tumour invasion; (E) rostrocaudal radiograph in which there is increased opacity affecting the right frontal sinus (arrow). This occurred more frequently in cats with nasal neoplasia, but is also observed in cats with rhinosinusitis (see text).
Of the various radiographic patterns of nasal disease described by Myer (1998), pattern 1 (normal radiographic appearance of both nasal passages) was associated with rhinitis and pattern 3 (areas of increased soft tissue opacity superimposed over areas of conchal destruction) was associated with nasal neoplasia (Table 6).
Classification of radiographic signs in 48 cats with rhinitis or nasal neoplasia
00023-8/unzip/10.1016_S1098-612X(03)00023-8-table6.png)
Based on scheme described by Myer (1998).
One cat was classified as pattern 1 (normal radiographic appearance of both nasal passages) despite presence of a metallic foreign body lodged just caudal to the external nares.
In each of the three cats with non-nasal disease in which the radiographs were interpreted as showing signs of intranasal disease, lesions affecting adjacent structures were superimposed on the nasal cavity in one or more of the radiographs evaluated. These included a cat with cutaneous lymphoma causing a swelling superimposed on the frontal sinuses, a cat with multiple fractures affecting the maxillae, nasal bones and mandible, and a cat with a large irregular exostosis centred on the alveolus of a canine tooth, which wassuperimposed on the adjacent nasal cavity, but which was considered to be an incidental, non-nasal condition on the basis of clinicalfindings.
Discussion
The cats in this study were referred for investigation of nasal signs of median duration of 3 months. We identified no examples of either cryptococcosis or aspergillosis, which are uncommon causes of rhinitis in cats in the UK (Goodallet al 1984, Coulson 1988, Gaskell 1994). The predominance of lymphoma and carcinomata in cats in this series with nasal neoplasia is compatible with previous reports. We identified no examples of squamous cell carcinoma, which is considered relatively prevalent but usually affects the nasal planum rather than intranasal structures (Levy and Ford 1994, Mukaratirwa et al 2001). One cat in this study had a nasal granular cell tumour, which has been reported rarely in the cat (Patnaik 1993).
Positioning for radiography and the radiographic anatomy of the feline skull have been described in some detail by Farrow (1994). All the cats in this study had a dorsoventral (intraoral) radiograph and nearly all had lateral and ventrodorsal or dorsoventral projections; however, there were variations in the number of radiographs obtained, which reflects attempts to optimise the radiographic examination according to the cats' clinical signs. For example, some cats also had a rostrocaudal radiograph to examine the frontal sinuses and some had oblique radiographs to examine the dental arcades. The projections that are probably most useful in dogs with nasal signs are the dorsoventral (intraoral) view of the nasal cavity and the rostrocaudal view of the frontal sinuses. These projections are usefulbecause there is minimal superimposition of the nasal cavity and frontal sinuses by other structures, and each provides an opportunity to compare left and right, which aids recognition of unilateral or asymmetrical lesions. It is difficult to obtain a satisfactory view of the frontal sinuses in rostrocaudal radiographs in breeds of cat with a short nose and relatively small frontal sinuses, such as Persians. The ventral 20° rostral-dorsocaudal oblique is a useful alternative projection for examining the frontal sinuses. Lateral radiographs are frequently unhelpful for assessing animals with unilateral lesions confined to the nasal cavity because superimposition of the normal half of the skull on the abnormal half tends to mask abnormalities; however, the lateral view aids identification of lesions extending through the nasal or frontal bones or destroying the cribriform plate (Fig 2D), which is a feature of nasal neoplasms that invade the brain (Smith et al 1989). Lateral radiographs are also useful for identifying lesions affecting the nasopharynx, such as polyps (Kapatkin et al 1990).
The difficulties of examining the feline nasal cavity radiographically are emphasised in this study by the finding that the radiographs of a few cats with non-nasal disease were interpreted erroneously as showing signs of intranasaldisease. This probably occurred because lesions affecting adjacent structures were superimposed on the nasal cavity in one or more of the radiographs evaluated. Computed X-ray tomography also appears to be a useful method for examining the nasal cavity and nasopharynx in cats (Losonsky et al 1997, Allen et al 1999) largely because it eliminates the problem of superimposition.
As has been reported previously (Cox et al 1991, Cape 1992, O'Brien et al 1996), we found considerable overlap in the radiographic signs in cats with rhinitis or nasal neoplasia. Of the various signs we assessed, displacement of midline structures, unilateral generalised soft tissue opacity, unilateral generalised loss of turbinate detail, evidence of bone invasion and unilateral nasal lesion(s) were significantly associated with nasal neoplasia, but the predictive value of these signs was relatively modest, in the range 56–73%. When examining radiographs of cats (or dogs) with nasal signs, finding a combination of signs increases the predictive value. For example, of the seven cats in this study that had a unilateral nasal lesion associated with generalised loss of turbinate detail, generalised increased soft tissue opacity and evidence of invasion of bones, six (86%) had nasal neoplasia. As described previously with respect to nasal disease in dogs (Russo et al 2000), differentiation of rhinitis and nasal neoplasia should ideally be based on finding combinations of signs that together have a high predictive value.
The radiographic signs we identified in cats with nasal neoplasia are similar to those reported in dogs (Sullivan et al 1987, Russo et al 2000). One noteworthy difference is the lower proportion of cats with nasal neoplasia that had recognisable fluid/soft tissue opacity in the frontal sinus (48%) compared to dogs (70%). This and the higher proportion of cats with rhinitis that also had fluid/soft tissue opacity in the ipsilateral frontal sinus (26%), combine to reduce the predictive value of this sign (Fig 2E), making it less useful as a means of distinguishing these conditions than it is in dogs. Frontal sinusitis is a common sequel to chronic rhinitis in cats (Bradley 1984, Cape 1992).
Radiographic signs in cats with rhinitis are more variable and less specific than those for nasal neoplasia, and in 30% affected cats in this study the radiographs appeared to be normal. The presence of lucent foci, which is an important sign of destructive rhinitis in dogs (Sullivan et al 1986, Russo et al 2000), was recognised in relatively few cats in this study. Lucent foci may be observed in cats with either foreign body (Bright 1981) or aspergillosis (Goodall et al 1984); however, these are uncommon diagnoses. The inclusion of four nasal foreign bodies in this study could be considered unrepresentative of their incidence in cats. Three of these were metallic airgun pellets, which greatly facilitated their detection. The remaining nasal foreign body, a piece of grass awn, was not visible radiographically.